From April-May 2020, Sentinel-Hub had organized the third edition of their custom script competition. The competition was organized in collaboration with the Copernicus EU Earth Observation programme, the European Space Agency and AI4EO consortium.

從2020年4月至5月，Sentinel-Hub組織了第三次自定義腳本競賽 。 該競賽是與哥白尼歐盟地球觀測計劃 ， 歐洲航天局和AI4EO財團合作組織的 。

Motive of the competition:

比賽動機：

To look for new, innovative scripts, which enable users to make sense of Earth Observation data. It was designed to find solutions to the huge challenges in the new ‘normal’

尋找新的，創新的腳本，使用戶能夠理解地球觀測數據。 它旨在為新的“ 正常 ”情況下的巨大挑戰找到解決方案

What was the objective of our submission:

我們提交的目標是什么 ：

Covid-19 has led to many world governments enforcing emergency quarantine measures. What is the impact of such policies on the environment? Can we measure the change in pollution levels? Is there a co-relation to economic activity? And can we build leading indicators which enable us to measure economic activity at a global scale. Through our project, we tried to analyze satellite imagery to identify the answers.

Covid-19導致許多世界政府強制執行緊急隔離措施。 此類政策對環境有何影響？ 我們可以衡量污染水平的變化嗎？ 與經濟活動是否有關聯？ 我們是否可以建立領先的指標，使我們能夠在全球范圍內衡量經濟活動。 通過我們的項目，我們嘗試分析衛星圖像以找出答案。

Tools used:

使用的工具：

The tools make the man, or so they say!

這些工具造就了男人，或者他們說！

I always tell my team to first start, by researching the available tools. Because the right tools can make or break a project!

我總是告訴我的團隊首先研究可用的工具。 因為正確的工具可以建立或破壞項目！

The tools that were available for this project were:

該項目可用的工具有：

Sentinel-Hub EO Browser

Sentinel-Hub EO瀏覽器

Sentinel-Hub API access

Sentinel-Hub API訪問

Picterra’s AI platform

Picterra的AI平臺

euroData Cube

euroData多維數據集

We first used Sentinel-5P data to plot NO2 levels. This was done to identify any anomalies (sudden changes) in the data. This was then collaborated with VHR imagery (Very high resolution imagery)

我們首先使用Sentinel-5P數據繪制NO2水平。 這樣做是為了識別數據中的任何異常(突然變化)。 然后與VHR影像(非常高分辨率的影像)合作

Statistical analysis of this data was used to show a significant difference in countries like Germany (where the lockdown was strictly implemented) vs countries like Italy.

通過對該數據進行統計分析，可以發現德國(嚴格執行鎖定)等國家與意大利等國家之間存在顯著差異。

Plot of NO2 levels for Germany & Italy (note the presence of missing values) Source: Sentinel-5P processed data德國和意大利的NO2水平圖(注意存在遺漏值)來源：Sentinel-5P處理的數據

I have uploaded the notebook used for plotting the above plot on euroData Cube’s Contribution page. you can check it via the link given below:

我已經將用于繪制上述圖表的筆記本上載到euroData Cube的“貢獻”頁面。 您可以通過以下鏈接進行檢查：

https://eurodatacube.com/marketplace/notebooks/contributions/NO2_Analysis_Covid19_Lockdowns.ipynb

https://eurodatacube.com/marketplace/notebooks/contributions/NO2_Analysis_Covid19_Lockdowns.ipynb

NO2 is produced by vehicular traffic and as un-burnt residue from chemical processes. This can be collaborated by counting cars from high resolution satellite imagery.

機動車交通產生的二氧化氮是化學過程中未燃燒的殘留物。 可以通過計算高分辨率衛星圖像中的汽車來進行協作。

We will now use Picterra’s machine learning platform to identify vehicles in VHR imagery.

現在，我們將使用Picterra的機器學習平臺來識別VHR圖像中的車輛。

We need to first login to Picterra’s geo-spatial imagery platform. They have a free trial version, which you can use (without providing any credit cards!)

我們需要先登錄Picterra的地理空間圖像平臺。 他們有一個免費的試用版，您可以使用它( 無需提供任何信用卡！ )

Once on the home page, you can Create a new project

進入主頁后，您可以創建一個新項目

Now you can upload images from your local drive (other options are Online map imagery OR the option to Buy satellite images). Once uploaded, you can then use it to train a detector.

現在，您可以從本地驅動器上載圖像 (其他選項是“ 在線地圖圖像”或“ 購買衛星圖像 ”選項)。 上傳后，您可以使用它來訓練探測器。

Once the images are uploaded, we then go to the training section. Here we can use the platform to train a detector. However there are pre-built detectors which you can use. But for this demo, we will train a new one

圖片上傳后，我們將轉到培訓部分。 在這里，我們可以使用平臺訓練探測器。 但是，您可以使用預制的探測器。 但是對于這個演示，我們將訓練一個新的

Training mode!訓練模式！ Available pre-trained models可用的預訓練模型

Click on the Train a new detector

單擊訓練新的檢測器

There are some custom base models. We chose the Vehicles option

有一些自定義基本模型。 我們選擇了車輛選項

We want to detect cars

我們要檢測汽車

There are two available detection type: i) Count and ii) Segmentation

有兩種可用的檢測類型：i)計數和ii)分段

We select the default Count option. and then press the Create button

我們選擇默認的“ 計數”選項。 然后按創建按鈕

Then add the images that you had added to the project. and press the Start Training button

然后添加您已添加到項目中的圖像。 然后按開始訓練按鈕

When you do it for the first time, Picterra gives you a nice video tutorial that walks you through the training process.

第一次進行此操作時，Picterra會為您提供一個不錯的視頻教程，引導您完成培訓過程。

The interface is very intuitive, and once you understand the controls, it becomes easy to navigate.

界面非常直觀，一旦您了解了控件，就可以輕松瀏覽。

First select the Training button to the left. After that you can select the area of interest, as marked by the yellow square. Clicking on the selected AOI, will zoom into the image

首先選擇左側的訓練按鈕。 之后，您可以選擇感興趣的區域，以黃色正方形標記。 單擊所選的AOI，將放大圖像

Now click on the Polygon button and select the Circle option

現在單擊“ 多邊形”按鈕，然后選擇“ 圓”選項

You can now click on each of the parked cars within the area of interest and draw a circle on top. This is now your annotated train dataset. You could have selected the polygon option and drawn a box across the car outline, but I am too lazy for that!

現在，您可以單擊感興趣區域內的每個停放的汽車，并在頂部繪制一個圓圈。 現在這是您帶注釋的火車數據集。 您可以選擇“多邊形”選項，并在汽車輪廓上繪制一個框，但是我實在太懶了！

Now select the Testing button on the left menu and click and mark the testing area on the image. You can annotate the testing dataset. But here I will skip that step — Just to check what happens!

現在，選擇左側菜單上的“ 測試”按鈕，然后單擊并在圖像上標記測試區域。 您可以注釋測試數據集。 但是在這里，我將跳過該步驟- 僅檢查發生了什么！

Now click on the Train Detector button at the top of the dashboard.

現在，單擊儀表板頂部的“ 火車檢測器”按鈕。

Now you can go and have a coffee ( I prefer my masala chai!)

現在您可以去喝咖啡了( 我更喜歡我的咖喱柴！ )

https://en.wikipedia.org/wiki/Masala_chai#/media/File:Masala_Chai.JPGhttps://en.wikipedia.org/wiki/Masala_chai#/media/File:Masala_Chai.JPG

OR/AND

或/與

can check the educational resources provided. Do check the Picterra University. It has some good resources on end-to-end machine learning for geo-spatial images.

可以查看提供的教育資源。 請檢查Picterra大學。 它在地理空間圖像的端到端機器學習方面有一些不錯的資源。

Since the training is now finished, let's go and check the performance of the detector.

由于培訓現已結束，因此我們開始檢查檢測器的性能。

Testing output測試輸出

The performance is good…but not great! At the bottom right, it has mis-identified the roof of the warehouse as a car.

表現不錯…但不是很好！ 在右下角，它錯誤地將倉庫的屋頂標識為汽車。

For a detector its important to know what it should not consider as an object. In machine learning terms - false positives should be low.

對于檢測器來說，重要的是要知道它不應該考慮什么。 用機器學習的術語來說，誤報率應該很低。

So we will annotate some more vehicles for our training set. We select a larger area near our previous training set.

因此，我們將為訓練集注釋更多的車輛。 我們在之前的訓練集附近選擇了一個較大的區域。

Post training, I get a warning: it seems I have some overlapping annotations in the training set. which is true. This can cause mergers while running the inference jobs and give an incorrect count.

訓練后，我得到一個警告：看來我在訓練集中有一些重疊的注釋。 沒錯 這可能會在運行推理作業時導致合并，并且計數不正確。

So let’s correct that -

所以讓我們更正-

Use the Select button on the left to select any of the annotations & modify the marker (green circle).

使用左側的“選擇”按鈕選擇任何注釋并修改標記( 綠色圓圈 )。

Modifying the training data to avoid merged annotations修改訓練數據以避免合并注釋

Retraining the detector, we get the following output.

重新訓練檢測器，我們得到以下輸出。

Testing area after re-training重新訓練后的測試區域

Seems good. We have missed one large white vehicle at the left, but now the misclassification from the roof is missing. And it has got the car to the extreme left corner.

看起來不錯。 我們錯過了左側的一輛大型白色車輛，但是現在缺少了從車頂分類錯誤的信息。 它已將汽車推到了最左端。

The detector can be trained with more data. Let’s now run the detector on the entire image to see the evaluation performance. We will also evaluate it on a new image.

可以用更多數據訓練檢測器。 現在讓我們在整個圖像上運行檢測器以查看評估性能。 我們還將在新圖像上對其進行評估。

Added new un-seen images & the trained detector添加了新的看不見的圖像和訓練有素的探測器

Press the Run Detector button in front of the image. A prompt informs you about the number of processing quota required for running the data. Since we are running on a free-quota, we press Start Detection.

按下圖像前面的“運行檢測器”按鈕。 提示會通知您有關運行數據所需的處理配額數量。 由于我們使用的是免費配額，因此請按開始檢測。

Prompt for the Processing quota for the detector提示檢測器的處理配額

Viewing the results shows us the output of the detector

查看結果可向我們顯示檢測器的輸出

Detector output from unseen images來自看不見圖像的檢測器輸出

We can also use this powerful platform, for other object detection activities, like solar-arrays, ships, trash, military vehicles etc.!

我們還可以使用這個功能強大的平臺進行其他物體檢測活動，例如太陽能電池板，輪船，垃圾箱，軍用車輛等！

Please check Picterra’s blog for their recent success stories:

請查看Picterra的博客，了解他們最近的成功案例：

• Object detection stories

  物體檢測故事

• Bring AI-powered object detection to ArcGIS

  將AI驅動的對象檢測帶到ArcGIS

Note on image licensees:

關于圖片被許可人的注意事項：

Unless mentioned otherwise, the author owns the licenses to the images.

除非另有說明，否則作者擁有這些圖像的許可證。

Indian mosiac (NO2 levels) was created by the author using Sentinel-Hub Platform

作者使用Sentinel-Hub平臺創建了印度洋霜(NO2含量)

Time series plots for No2 levels in Germany & Italy were created by the author using code specified above

作者使用上面指定的代碼創建了德國和意大利2級水平的時間序列圖

Screenshots are from Picterra’s geo-spatial platform

屏幕截圖來自Picterra的地理空間平臺

翻譯自: https://towardsdatascience.com/how-i-won-sentinel-hub-covid-19-custom-script-hackathon-be882ed05186

總結

以上是生活随笔為你收集整理的如何使用Picterra的地理空间平台分析卫星图像的全部內容，希望文章能夠幫你解決所遇到的問題。

如果覺得生活随笔網站內容還不錯，歡迎將生活随笔推薦給好友。